**** BEGIN LOGGING AT Thu Mar 21 02:59:58 2013 Mar 21 03:01:54 guessing i would need to send i2c UBX configuration info at startup Mar 21 03:02:18 MrCurious: If it doesn't have a lock, you probably won't get much for a few minutes. Mar 21 03:02:36 MrCurious: Though you can send it data like the time/date to help it out. That's in the protocol manual Mar 21 03:02:40 in nmea mode it spams useless data for a few minutes.... Mar 21 03:02:48 yeah Mar 21 03:02:50 perhaps i2c mode is quieter until it locks Mar 21 03:03:11 I2C mode shouldn't send anything unless you ask for it. Mar 21 03:04:12 i think i am forseeing more document reading in my future :D Mar 21 03:04:20 MrCurious: Is this your own board, or one of those ebay specials? Mar 21 03:05:33 got it off amazon. has gnd +5 +3.3 rxd txd scl sda Mar 21 03:05:54 has a ublox neo 6m on it Mar 21 03:06:47 could also be that i am not as good at i2c in linux as i think i am :D Mar 21 03:08:38 MrCurious: I dunno, I've never I2Cd in linux Mar 21 03:08:57 only ever i2cd on arduinos Mar 21 03:10:55 MrCurious: looks liek you'd just i2cset BUS_ADDR CHIP_ADDR DATA_ADDR Mar 21 03:11:13 MrCurious: And then i2cget BUS_ADDR CHIP_ADDR Mar 21 03:11:19 to read its response Mar 21 03:12:29 like I said, I've yet to try it ;) Mar 21 03:12:46 I'd probably just stick with the serial port on the 'bone, TBH Mar 21 03:12:57 I just brought out the i2c lines because I could. Mar 21 03:20:15 sent Mar 21 03:20:22 order for 3000 pcbs Mar 21 03:23:05 win Mar 21 03:23:56 aaaaand updated SD card Cape http://i37.photobucket.com/albums/e94/MasterMWL/SDCape_zps8e6d37c4.png Mar 21 03:33:43 hmmmmmm Mar 21 03:34:26 no, but if you hum a few bars... Mar 21 03:37:26 https://fbcdn-sphotos-h-a.akamaihd.net/hphotos-ak-prn1/p480x480/45377_10151513354867661_1614812827_n.jpg Mar 21 03:37:30 this is todays effort Mar 21 03:41:14 jkridner: "Dog Tick Board" - http://www.petparents.com/content/images/icon_tick.gif Mar 21 03:41:30 sounds painful Mar 21 03:41:34 hehe Mar 21 03:41:57 Shadyman: ooh, you selling those SD card capes? Mar 21 03:42:13 jkridner: Eventually, yes Mar 21 03:42:29 gps too for that matter. Mar 21 03:42:44 and parallel port cape ;) Mar 21 03:42:54 yes, a parallel port for the bbone :/ Mar 21 03:43:44 mostly because I found the 74LVX161284, and it would be a shame not to use it. Mar 21 03:44:02 Shadyman: hey, for your "$10 uBlox", what volumes are you talking about? Mar 21 03:44:08 ds2: 1. Mar 21 03:44:17 Shadyman: new? Mar 21 03:44:19 yes. Mar 21 03:44:31 hmmm ok Mar 21 03:44:47 but like I said, I've yet to actually test it out :) Mar 21 03:44:54 clearly we looking at different sources Mar 21 03:44:58 yeah. Mar 21 03:45:03 all my sources show > $10 for uBlox Mar 21 03:45:11 others are cheaper Mar 21 03:45:19 oh? Mar 21 03:45:22 but this is for higher quantity then 1 Mar 21 03:45:46 you've piqued my interest Mar 21 03:46:11 uBlox store sells them for about $90 each Mar 21 03:46:36 ds2: ublox store is insanely overpriced. Mar 21 03:46:50 I think it's cheaper once you sign up for a contract account or something Mar 21 03:47:01 what is a ublox Mar 21 03:47:27 Shadyman: this is the no Antenna version, right? Mar 21 03:47:48 oh gps stuff Mar 21 03:47:53 mrpackethead: yeah Mar 21 03:48:06 ds2: It needs an antenna Mar 21 03:48:12 ah Mar 21 03:48:16 ds2: chip, external, whichever Mar 21 03:48:18 mrpackethead: I got an error on http://www.stellascapes.com/ trying to send a message: SMTP Error! Could not connect to SMTP host Mar 21 03:48:37 oh.. Mar 21 03:48:39 just email me Mar 21 03:48:45 andrew.frazer@stellascapes.com Mar 21 03:48:59 our web page is hiddiolusly out of date Mar 21 03:49:03 if we got it undercontrol Mar 21 03:49:12 yikes what woudl heappen Mar 21 03:56:25 hell would freeze over? Mar 21 04:00:51 ;) Mar 21 05:05:04 Shadyman: hell doesn't exist Mar 21 05:05:15 * Shadyman shrugs Mar 21 05:05:22 but you'd certainly cool it down ;) Mar 21 05:05:30 mranostay: thats what the devil wants you to belive Mar 21 05:06:43 mrpackethead: hell ... is other people Mar 21 05:06:44 * mranostay wants to violently stab people that only have parts in brd files and not in a library Mar 21 05:06:57 lol. Mar 21 05:07:04 mranostay: you can convert Mar 21 05:07:13 theres a ulp for that Mar 21 05:07:28 yes i know but christ why should i have too? Mar 21 05:07:41 because you dont' normally send all your library files with a board file Mar 21 05:07:42 ? Mar 21 05:08:09 well true but you'd think you'd post your changes to a libary if you had them on github Mar 21 05:08:11 i cna't understand your problem. Mar 21 05:08:23 maybe. Mar 21 05:08:24 i cna't understand ur speeling Mar 21 05:08:38 wyh Mar 21 05:08:39 ? Mar 21 05:08:44 bcuz Mar 21 05:08:50 mranostay: eagle? Mar 21 05:08:56 yes Mar 21 05:09:06 eagle 6 or eagle 5? Mar 21 05:09:21 there is a ulp to generate the library from a brd file Mar 21 05:09:27 or you can just cut and paste the part Mar 21 05:10:59 i need to learn to use GIT Mar 21 05:11:05 lol. Mar 21 05:11:13 maybe thats todays job Mar 21 05:11:14 :-) Mar 21 05:11:16 it is just one command.... how hard can it be? Mar 21 05:29:51 that is more than a day :P Mar 21 05:30:51 git is a 15minute task Mar 21 05:36:54 initially Mar 21 06:09:24 so, how do you merge none text based files in git? Mar 21 06:12:55 mrpackethead, --ours/--theirs? Mar 21 06:13:18 you can add external merge tools Mar 21 06:14:33 like eagle pcb files Mar 21 06:14:34 :-) Mar 21 06:14:39 sure Mar 21 06:14:47 i've just installed gitlab Mar 21 06:14:52 on our server Mar 21 06:14:58 now gotta fihure out how to use it Mar 21 06:14:59 :-) Mar 21 06:18:37 * mranostay review his algorithmns pdf Mar 21 06:19:52 prpplague: what is gitlab exactly? Mar 21 06:20:51 er mrpackethead Mar 21 06:21:16 http://blog.gitlab.com Mar 21 06:22:17 private git server Mar 21 06:22:26 was'nt too keen to put everything on github Mar 21 06:22:33 we have no way of controlling it Mar 21 06:22:36 i swear the only time i use sorts and searchs are in job interviews ... Mar 21 06:26:17 I love the xkcd job interview sort Mar 21 06:26:28 what is sort? Mar 21 06:26:36 sort algorithm Mar 21 06:26:52 http://xkcd.com/1185/ Mar 21 06:30:43 Russ: if i see the interview going badly i'll do the panic sort :P Mar 21 06:31:50 in case you didn't see it: http://gkoberger.github.com/stacksort/ Mar 21 06:32:16 i i created my project Mar 21 06:32:22 wonder how i move it onto the server Mar 21 06:32:23 :-) Mar 21 06:37:09 heh Mar 21 06:40:03 git is not a client server setup Mar 21 06:54:31 so, its sort of two servers Mar 21 06:54:32 :-) Mar 21 06:54:38 and two clients Mar 21 06:54:45 depends which way you look at it Mar 21 06:57:58 it's a DVCS Mar 21 07:02:13 git can be just a single client Mar 21 07:02:23 there is no need for more then 1 of anything Mar 21 07:15:25 it can be just 1 Mar 21 07:15:27 or 2 Mar 21 07:15:29 or 21 Mar 21 07:15:39 i think thats what makes it nice. Mar 21 07:15:40 :-) Mar 21 07:15:56 however having a common "server" where eveyone comes back to is kinda useful Mar 21 07:16:23 i'll use "server" in the loosest sense of the word Mar 21 07:23:06 grr will i ever sleep at a normal hour? Mar 21 07:24:44 mranostay: No. Mar 21 07:25:13 Shadyman: yes that is correct Mar 21 07:26:21 gm Mar 21 07:27:17 woglinde: hey old due Mar 21 07:27:19 *due Mar 21 07:27:22 gahahah Mar 21 07:27:25 *dude Mar 21 07:28:44 mranostay: I finally got to bed around noon Mar 21 07:29:39 hopefully today will be better :/ Mar 21 07:30:43 <_av500_> yarn Mar 21 07:31:12 ok _av500_ is here Mar 21 07:31:19 now it's a party. Mar 21 07:31:20 good enough reason to sleep Mar 21 07:31:43 oh Mar 21 07:31:44 fine then Mar 21 07:32:00 Shadyman: it isn't you it is _av500_ Mar 21 07:32:05 <_av500_> mranostay: have some nightmares Mar 21 07:32:10 mranostay: Sure, you say that now Mar 21 07:32:25 * mranostay hugs _av500_ and Shadyman goodnight Mar 21 07:32:29 bai Mar 21 07:32:31 <_av500_> Shadyman: he will do the same to you one day Mar 21 07:32:36 which one of you is reading me a nighttime story? Mar 21 07:32:44 <_av500_> I can cat you a TRM Mar 21 07:32:51 <_av500_> pastebin? Mar 21 07:33:04 can you do TRM in Dr Suess style? Mar 21 07:33:19 <_av500_> one of the PMIC ones then Mar 21 07:33:26 Shadyman: btw nice quote Mar 21 07:33:35 er? Mar 21 07:33:42 http://26-26-54.hardwarebug.org/152 Mar 21 07:33:45 <_av500_> need to drag a kid to school, bbl Mar 21 07:33:53 ah Mar 21 07:40:10 hoi zme Mar 21 07:42:08 Shadyman: image file does not exist anymore Mar 21 07:43:09 KotH: http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape_zps965dc530.png Mar 21 07:44:13 Shadyman: and the schematics? :) Mar 21 07:44:26 working on it :/ silly caching webserver Mar 21 07:44:30 Shadyman: where are the C1's for the RF part? are they on the bottom? Mar 21 07:44:46 ah.. found them Mar 21 07:45:35 KotH: http://erroraccessdenied.com/GPS_Cape2.png Mar 21 07:45:47 boo small Mar 21 07:46:48 Shadyman: move U4+C8 to the right, so you can move the power trace also to the write, then move the via that connects R19 so that C7's GND goes directly to the plane instead of over the contact of C6 Mar 21 07:47:17 Shadyman: also, what width is that power trace? 0.3mm? Mar 21 07:48:48 10 mil Mar 21 07:49:07 hi cheeae Mar 21 07:50:21 KotH: Gave C7 its own finger to ground Mar 21 07:52:10 KotH: Woudl be easy enough to make the power line 25 mil Mar 21 07:52:28 hello! I'm working with a project on the beaglebone, and we would like to use a 1C lipo on the P6 battery header. Our battery does not have NTC, and I've come to understand that we have to emulate the temperature sensing. Does anyone have some good sources as to how to connect the battery? Also, what is the purpose of the pin "Battery sense" ? Any help is much appriechiated :) Mar 21 07:53:04 Shadyman: additional width is always desirable, but over 0.3mm gets kind of pointless in this application (the neo doesnt need too much peak.. iirc it's 80mA or something) Mar 21 07:53:20 155mW at 3V, yeah Mar 21 07:53:24 kristian_: soldering iron works sweet Mar 21 07:53:25 :) Mar 21 07:53:34 KotH: That's why I didn't bump it up bigger Mar 21 07:53:52 Shadyman: how does the buttom layer look? any larger cuts there? Mar 21 07:54:02 Shadyman: especially those going into the 6cm range? Mar 21 07:54:31 KotH: They're all via-spammed. Lemme get a pic Mar 21 07:54:49 mrpackethead: hehe, well given some schematic I can solder anything :p Mar 21 07:54:50 Shadyman: also, did you check what is underneath the antenna connector on the beagle? Mar 21 07:55:30 KotH: That's where the GPIO LEDs are Mar 21 07:59:10 gm koth Mar 21 07:59:24 koth no courses today? Mar 21 07:59:26 hoi woglinde_ Mar 21 07:59:27 KotH: http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape2_zpsc62b4950.png Mar 21 07:59:36 woglinde_: nope, only mon-wed Mar 21 07:59:59 Shadyman: and the usb hub Mar 21 08:00:16 Shadyman: will you use a standard standing sma connector Mar 21 08:00:28 KotH: right-angle, yes Mar 21 08:00:46 Shadyman: if so, you have to make sure not to push it all the way down, otherwise you short the antenna trace with the shield ring of the connector Mar 21 08:01:02 Shadyman: maybe a edge sma connector would be better in this case Mar 21 08:01:35 Short the antenna trace with the shield ring? How does that happen? Mar 21 08:02:31 the shield goes directly over the trace, forming a small capacitance Mar 21 08:04:08 huh. Mar 21 08:05:11 hehe Mar 21 08:05:16 i did this mistake as well :) Mar 21 08:05:24 didnt see it until i had soldered the connector on Mar 21 08:05:40 even though it's soldermasked? Mar 21 08:06:34 so how are you supposed to feed an sma, then? From another layer? Mar 21 08:06:39 the solder mask is only 0.01-0.1mm you get a very low ESR 1-2pF capacitor :) Mar 21 08:07:07 best is either to come from the opposite side, or to use an sma edge connector Mar 21 08:08:08 the sma drawing I have suggests that it stands off the board by 0.4mm Mar 21 08:08:31 http://www.te.com/commerce/DocumentDelivery/DDEController?Action=srchrtrv&DocNm=SMA-JR-P2&DocType=Customer+Drawing&DocLang=English&PartCntxt=2081235-1 Mar 21 08:09:58 ah.. this has standoffs... then it's ok Mar 21 08:10:11 phew, had me scared there :/ Mar 21 08:10:15 http://ch.mouser.com/ProductDetail/Linx-Technologies/CONSMA003031/?qs=sGAEpiMZZMuLQf%252bEuFsOru48dlYVRPu%252bGmcmDuihAj8%3d Mar 21 08:10:24 this would be an edge connector Mar 21 08:10:52 I've never seen a non-sloppily-mounted edge connector :/ Mar 21 08:11:08 dont let them solder in china :) Mar 21 08:11:12 heh. Mar 21 08:11:37 on the bottom side, where the 4 traces come together, you want to move them apart, so the ground flows around them Mar 21 08:11:54 otherwise you have very long cuts in the ground plane which acts as a high performance antanna Mar 21 08:11:58 antenna* Mar 21 08:12:22 even with all the vias? Mar 21 08:13:17 i mean, it's doable, i just wanted to stay as far away from the rf section as possible :) Mar 21 08:13:54 yes Mar 21 08:14:02 google for slit antenna Mar 21 08:15:17 this is why some of us just buy ready made modules when it comes to RF stuff Mar 21 08:15:18 :-) Mar 21 08:15:24 mrpackethead: hehe Mar 21 08:15:27 becuase its hard Mar 21 08:15:29 :-) Mar 21 08:23:58 KotH: http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape2_zpseddee2fa.png Mar 21 08:24:11 I'll have to disable the ground flood to get at that orphan trace Mar 21 08:24:18 but it's on the list Mar 21 08:25:40 Antenna theory, what did I ever do to you?? Mar 21 08:26:02 this thing is killing me Mar 21 08:26:06 how hard can it be Mar 21 08:26:28 mrpackethead: Which thing? Mar 21 08:26:35 git Mar 21 08:26:38 ah Mar 21 08:26:40 heh Mar 21 08:26:46 git clone is as far as I go Mar 21 08:27:05 yeah, i've done that Mar 21 08:28:04 might be working now Mar 21 08:28:41 hey hey! Mar 21 08:30:02 treat git nicely and it will treat you nicely back Mar 21 08:30:22 re av500 Mar 21 08:30:27 siberia Mar 21 08:30:32 never ends Mar 21 08:31:16 mrpackethead: http://i37.photobucket.com/albums/e94/MasterMWL/308402_200894596651257_126894987384552_458333_1459949569_n.jpg Mar 21 08:31:34 Shadyman: dont forget the two vias that connect the battery power and the baseband power to the top layer Mar 21 08:31:44 Shadyman: these form the longest slit Mar 21 08:31:50 lol. Mar 21 08:32:16 mrpackethead: RF is not hard, you just have be more carefull what you are doing Mar 21 08:32:29 woglinde: it just started snowing here Mar 21 08:32:39 things become relevant that you can ignore with DC stuff Mar 21 08:32:41 many things are not hard, when you have experiecne and know-how Mar 21 08:32:54 mrpackethead: and keep in mind, modern uC/uP circuits are in the RF category already Mar 21 08:33:01 without those, things are hard Mar 21 08:33:02 here it snows since days Mar 21 08:33:09 including RF Mar 21 08:33:19 KotH: but I miss the tuning pots on them Mar 21 08:33:44 mrpackethead: know-how is easy to get by, just pick up a good book on rf stuff in your friendly library Mar 21 08:34:01 mrpackethead: or talk with your amateur radio friends Mar 21 08:34:04 av500: hehe Mar 21 08:34:15 av500: we could make a "pot-cape" for you :) Mar 21 08:34:20 the internet killed all the ham guys Mar 21 08:34:35 mrpackethead: ham guy next to me is quite alive Mar 21 08:34:55 the two 2 doors down 2 Mar 21 08:34:57 too Mar 21 08:34:58 mrpackethead: dm8tbr and ka6sox also seem to be alive Mar 21 08:35:12 * dm8tbr looks Mar 21 08:35:16 KotH: you need some shades of grey in your world Mar 21 08:35:17 :-) Mar 21 08:35:32 * KotH doesnt get that reference Mar 21 08:35:36 rumor has it that we hams even use the intarwebz... Mar 21 08:35:53 also 44./8 and all that Mar 21 08:37:03 * mrpackethead also has a radio operations licence Mar 21 08:37:48 I *think* its limited to some speical use cases Mar 21 08:37:52 * av500 does not have one Mar 21 08:38:00 * av500 is surrounded by hams Mar 21 08:38:10 * av500 would rather be surrounded by ham Mar 21 08:38:38 ham and other processed meats have been shown to increase your risk of cardio vascular issues by 70% Mar 21 08:38:42 * KotH would rather be surrounded by chocolate Mar 21 08:38:44 i've given up on bacon Mar 21 08:39:20 bacon is not that much "processed" Mar 21 08:39:24 just salty and dry Mar 21 08:39:29 and fatty Mar 21 08:39:50 unless you buy that cheap usian shit that has been treated with more chemicals then it's dry weight Mar 21 08:40:23 KotH: Do I need to worry about slit antennas on the baseband side? Mar 21 08:40:25 http://www.medicalnewstoday.com/articles/257362.php Mar 21 08:40:42 Shadyman: probably.. but do you eat bacpn.. that is much more to be worried about Mar 21 08:40:53 Shadyman: worry about slit-thoraits Mar 21 08:40:58 Shadyman: worry about slit-throats Mar 21 08:41:00 darn Mar 21 08:41:20 http://www.medicalnewstoday.com/images/articles/257362-bacon-bap.jpg Mar 21 08:41:47 that! Mar 21 08:41:52 [kernel] koenkooi pushed 1 new commit to 3.8: http://git.io/lDdSRQ Mar 21 08:41:52 kernel/3.8 1d68ca6 Koen Kooi: 3.8: update to 3.8.4... Mar 21 08:42:02 URGGHHGHGH!!! Mar 21 08:42:03 mrpackethead: You leave backbacon out of this Mar 21 08:42:42 basically it says " eat your vegetables " Mar 21 08:42:55 BegleGitbhub is a bot Mar 21 08:43:01 its sucks Mar 21 08:43:17 my other channel #openlighting, has a no bots rule! Mar 21 08:43:22 Kernel panic - not syncing: No init found. Try passing init= option to kernel. See Linux Documentation/init.txt for guidance. Mar 21 08:43:23 you bot, i'll bot you Mar 21 08:43:51 KotH: http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape2_zpsaf7aebc3.png Mar 21 08:44:51 anyone can understand chinese Mar 21 08:45:16 billions of people Mar 21 08:45:52 tw___: 我有点明白了中国 Mar 21 08:46:35 haha Mar 21 08:46:52 mrpackethead: Your mandarin brings all the boys to the yard Mar 21 08:47:21 hello everyone Mar 21 08:47:33 请停在超市里的蛇 Mar 21 08:47:41 panic__: ahhh! why. Mar 21 08:49:04 http://blog.csdn.net/cpuabc/article/details/7796127 i used this page to build MLO and boot.bin Mar 21 08:51:30 and used https://github.com/Aptina/BeagleBoard-xM/blob/master/MT9V034/Angstrom/README_Beagleboard-xM_mt9v034.txt this (kernel source linux3.1.2 to build uImage, then kernel panic. !! Mar 21 08:52:34 panic__: if you want a working mt9v034, just follow the angstrom instructions, that will get you a bootloader and kernel with that Mar 21 08:55:15 whois patina? Mar 21 08:55:22 I mean aptina Mar 21 08:55:24 /whois Mar 21 08:55:30 woglinde: whois google Mar 21 08:55:49 av50 thanks Mar 21 08:55:58 do not know what steps error Mar 21 08:57:25 how to build u-boot and kernel in beagleboard xm Mar 21 08:58:00 panic__: www.angstrom-distribution.org/building-angstrom Mar 21 08:58:03 install a compiler on your board Mar 21 08:58:22 koen you are covering native builds now? Mar 21 08:58:52 only naive builds Mar 21 08:59:21 woglinde: I try to make native builds work, but 99% of what I use is crosscompiled Mar 21 09:00:06 i have been install compile tooltrain Mar 21 09:00:55 09:57 < koen> panic__: www.angstrom-distribution.org/building-angstrom Mar 21 09:01:25 time flies Mar 21 09:01:45 tempus fluit, hora ruit Mar 21 09:02:24 day ruint Mar 21 09:09:20 well, we are getting there Mar 21 09:09:21 :-) Mar 21 09:09:28 KotH: http://i37.photobucket.com/albums/e94/MasterMWL/GPS_Cape_zps6cb947cb.png I separated the data lines from the baseband as well. Mar 21 09:09:56 is that cape swiss? Mar 21 09:10:06 mrpackethead: lol Mar 21 09:10:15 mrpackethead: I prefer Havarti. Mar 21 09:10:41 im am on my 3rd glass of Merlot Mar 21 09:12:18 mrpackethead: Lightweight. Mar 21 09:16:15 Shadyman: you've to worry about slit antennas everywhere, when you have one component that goes into the GHz range Mar 21 09:16:31 Shadyman: and that can be even one LVX logic gate Mar 21 09:16:32 like bitbanged gpio Mar 21 09:17:32 Shadyman: looks ok Mar 21 09:22:46 have anyone any x-loader source and kernel source ? i want to build MLO, boot,bin and uImage in beagleboard xm Mar 21 09:22:58 www.angstrom-distribution.org/building-angstrom Mar 21 09:23:41 KotH: Woot Mar 21 09:24:01 now, label all vias in the silk screen Mar 21 09:24:09 Shadyman: does the timepulse pin go to an input pin where you can raise an irq? Mar 21 09:24:39 KotH: It goes to an eCCP(?) pin Mar 21 09:24:43 Shadyman: if you say yes and build the thing with a t model, you'll have all the time nuts going for that board Mar 21 09:24:59 * KotH doesnt have the TRM open Mar 21 09:25:44 Hello! Anyone here used the battery header (P6) on the beaglebone? Mar 21 09:25:52 KotH: Oh, eCAP. Mar 21 09:25:55 that's what TI calls it Mar 21 09:26:34 lemme pull up the trm Mar 21 09:27:00 Shadyman: best would be if it goes to a pin where you have access to a capture compare unit Mar 21 09:27:15 Shadyman: or a PRU or both :) Mar 21 09:27:23 Both. Mar 21 09:27:32 PRU-ICSS includes one eCAP Mar 21 09:28:43 KotH: http://pastebin.com/XUncDtxV Mar 21 09:28:59 that's jkridner's pastebin from yesterday :) Mar 21 09:33:10 that pin can either be set as a PRU-ICSS eCAP or a regular CPU eCAP Mar 21 09:35:02 I've yet to figure out exactly what that entails, but jkridner says it's better than an average gpio interrupt Mar 21 09:36:45 well it can do fancy capture stufdf Mar 21 09:36:52 like count how many times it was triggered Mar 21 09:36:58 or count how long between triggers Mar 21 09:37:17 so when you get the IRQ, you just read the counter values Mar 21 09:38:17 robert have me a pile of useful info on creating debian systems Mar 21 09:41:32 KotH: It seems I should run a line for EXTINT in that case. Mar 21 09:42:17 though the timepulse is great, apparently the extint on a 6t can be used to synchronize it. I'll have to l ook into it more in the morning Mar 21 09:42:52 g'nite :) Mar 21 09:44:25 ig'nite! Mar 21 09:45:28 * Shadyman ignites. Mar 21 09:49:00 ShadowJK: the extint line is meant as to timestamp signals with the u-blox module Mar 21 09:49:11 ShadowJK: some people want to do that, but usually it's the other way round Mar 21 09:49:21 er.. Mar 21 09:49:23 wrong guy Mar 21 09:49:31 why do people go offline for sleeping? Mar 21 09:50:00 no idea Mar 21 09:53:39 no porxy like me Mar 21 09:54:41 KotH, fear of typing in their sleep Mar 21 09:55:54 wasnt there an xkcd about that? Mar 21 10:45:59 ...U-Boot expects to find boot.scr.uimg... Mar 21 10:46:01 uh? Mar 21 10:46:16 heh Mar 21 10:46:30 which u-boot target? Mar 21 10:46:38 I cannot say it aloud Mar 21 10:46:46 but its a fruit Mar 21 10:47:16 Macintosh? Mar 21 10:47:22 Blackbery? Mar 21 10:47:40 latter is close Mar 21 10:48:09 treachery abounds Mar 21 10:48:32 see bradfa's g+ post... Mar 21 10:49:10 which is more likely to be broken, rs232 or hdmi driver? Mar 21 10:50:21 yeah. I've followed those patches...surprised swarren prefers boot.scr-ery Mar 21 10:50:43 well, he's at nvidia Mar 21 10:50:50 mru, on what day of the week? Mar 21 10:50:59 mdp: caturday Mar 21 10:51:02 mru, yes... Mar 21 10:51:10 hdmi Mar 21 10:51:36 mru: you dont understand Mar 21 10:51:44 poor uk kids have access to family flatscreen Mar 21 10:51:47 but not to RS232 Mar 21 10:52:07 mru: and where would they hook up RS232 too? Mar 21 10:52:14 to the laptop they dont own Mar 21 10:52:17 that's true in .us Mar 21 10:52:18 hence the rpi Mar 21 10:52:32 very funny Mar 21 10:52:55 well, the premise of rpi is that kids w/o computer can learn computer Mar 21 10:53:23 voldemort board even figured out which interface was most important Mar 21 10:53:24 it's not to shift old stock of chips at minimum loss? Mar 21 10:53:37 HERESY! Mar 21 10:53:52 cynic Mar 21 10:54:03 CYRESY! Mar 21 10:54:14 maybe I'm a cynic, but I'm usually right Mar 21 10:54:25 nobody likes a smartarse :) Mar 21 10:55:00 we used Lego to reduce our old stock of chips Mar 21 10:55:13 won't be as good Mar 21 10:55:23 mdp: had LEGO uses BCM, the EV3 could play 1080p with XBMC Mar 21 10:55:27 used* Mar 21 10:55:33 heh Mar 21 10:56:04 instead of a monochrome 178x128 spi display Mar 21 11:19:02 damn, need to reboot Mar 21 11:19:17 someone install windows on your machine? Mar 21 11:19:21 nah Mar 21 11:19:27 it behaves strangely Mar 21 11:19:38 and I have 2 root shells stuck in "sync" Mar 21 11:20:07 and binaries started segfaulting that worked a few days ago Mar 21 11:20:22 dont laugh Mar 21 11:20:25 check your disks Mar 21 11:20:36 that will take the longest time of the reboot Mar 21 11:20:39 :) Mar 21 11:20:44 you need to reboot for that? Mar 21 11:20:58 no Mar 21 11:21:04 but I need to reboot anyway Mar 21 11:21:15 to shuffle some disks Mar 21 11:21:23 so, its the time Mar 21 11:21:23 those stucks processes, are they in D state? Mar 21 11:21:26 -s Mar 21 11:21:44 root 12933 0.0 0.0 2048 516 pts/2 D+ Feb21 0:00 sync Mar 21 11:21:53 D indeed Mar 21 11:21:58 D+ even Mar 21 11:22:03 I prefer G+ Mar 21 11:22:17 bad hw? Mar 21 11:22:23 no idea Mar 21 11:22:37 also cc1[23101]: segfault at ffffffc3 ip 0827aae3 sp bfca68cc error 4 in cc1[8048000+c5c000] Mar 21 11:22:41 in dmesg Mar 21 11:23:00 as I said, check the disks Mar 21 11:23:02 and ram Mar 21 11:23:12 bbl :) Mar 21 11:23:14 yeah.. first guess would be dead hd Mar 21 11:23:27 second dead ram Mar 21 11:23:30 third dead mainboard Mar 21 11:23:44 forth: not enough chocolate Mar 21 11:23:45 that's why you use raid and hotswap them Mar 21 11:36:15 mru: can you give me a hand? i'm looking for the alignment requirements of structs in the c99 standard.. but i'm unable to find it Mar 21 11:36:30 because it's not there Mar 21 11:36:37 oh.. Mar 21 11:36:39 it's in the abi specs of your target system Mar 21 11:36:44 ok, that explains it Mar 21 11:36:46 and where is it? Mar 21 11:37:12 c99 just says "things might need to be aligned" Mar 21 11:37:27 oki, thanks Mar 21 11:38:11 what's your target? Mar 21 12:02:29 fun Mar 21 12:02:40 mru: my segfaulting cc1 works again Mar 21 12:02:56 and the disks? Mar 21 12:03:01 seems fine Mar 21 12:03:08 check the smart log Mar 21 12:03:08 I was out for lunch, bit it booted Mar 21 12:03:12 yeap Mar 21 12:03:38 but yes, havent sent back seagates in a while Mar 21 12:03:38 and replace it if there's _any_ error Mar 21 12:03:42 yes Mar 21 12:04:00 I have a seagate waiting to be sent back Mar 21 12:04:41 I replaced 3 seagates last year Mar 21 12:05:08 I'm not against their service Mar 21 12:05:20 use the disk for 2.5ys, then get a new one Mar 21 12:05:28 I just wonder if that pays for them :) Mar 21 12:06:12 datapoint: don't mount emmc with -o discard Mar 21 12:07:34 this was one of the last with 5y warranty Mar 21 12:07:40 expires in june Mar 21 12:07:49 their timer was running a little fast Mar 21 12:08:01 what should I look at in smart report? Mar 21 12:08:09 SMART Error Log Version: 1 Mar 21 12:08:10 No Errors Logged Mar 21 12:08:21 any reallocated sectors? Mar 21 12:08:35 5 Reallocated_Sector_Ct 0x0033 100 100 036 Pre-fail Always - 0 Mar 21 12:09:05 I read that as 0 Mar 21 12:09:07 seems fine Mar 21 12:09:08 non? Mar 21 12:09:10 yeah Mar 21 12:09:26 4 Start_Stop_Count 0x0032 100 100 020 Old_age Always - 13 Mar 21 12:09:29 13 :) Mar 21 12:09:37 pretty good Mar 21 12:09:51 9 Power_On_Hours 0x0032 075 075 000 Old_age Always - 22584 Mar 21 12:10:12 9 Power_On_Hours 0x0032 093 093 000 Old_age Always - 36149 Mar 21 12:13:37 hello! I'm working with a project on the beaglebone, and we would like to use a 1C lipo on the P6 battery header. Our battery does not have NTC, and I've come to understand that we have to emulate the temperature sensing. Does anyone have some good sources as to how to connect the battery? Also, what is the purpose of the pin "Battery sense" ? Any help is much appriechiated :) Mar 21 12:16:15 mru: the same as always: msp430 Mar 21 12:16:22 av500: btw, why do your tablets randomly switch off every few days? Mar 21 12:16:31 battery is nowhere near empty Mar 21 12:16:55 which one? Mar 21 12:17:40 mru: i guess it's at least byte allingnet as the msp can do non-alignet read/write Mar 21 12:18:07 kristian_: just connect a resistor instead of the NTC Mar 21 12:18:38 av500: g9 Mar 21 12:19:20 KotH: check the compiler documentation Mar 21 12:19:27 juup Mar 21 12:19:29 on it Mar 21 12:19:38 I doubt there are any standard abis for msp430 Mar 21 12:19:46 juup Mar 21 12:19:52 mru: lets say that battery charging was a bigger challenge for TI than the whole OMAP4 design Mar 21 12:20:26 mru: speaking of ABI Mar 21 12:20:43 mru: is there a way to tell GCC to e.g. use only r0-r7 or so? Mar 21 12:20:55 or rather r8-r15 Mar 21 12:20:55 why would you want to do that? Mar 21 12:21:03 FIQ code Mar 21 12:21:09 I was about to ask that Mar 21 12:21:34 but I guess I would need to write a new ABI Mar 21 12:22:43 if I wanted function calls in my FIQ code Mar 21 12:29:13 av500 please use pastebin ;) Mar 21 12:29:52 sigh, I didn't pay enough attention when I bought the replacement cover glass and touchscreen for my phone. I bought one without cover frame. looks like that are the only two parts glued together Mar 21 12:29:59 * janne gets the hair dryer Mar 21 12:30:50 woglinde: for a single line? Mar 21 12:31:16 ah hm Mar 21 12:32:16 av500: isnt there any FIQ code attribute in gcc? Mar 21 12:32:43 hmm Mar 21 12:38:47 there is Mar 21 12:38:50 no idea what it does Mar 21 12:38:55 manual doesn't say Mar 21 12:39:30 http://gcc.gnu.org/ml/gcc-patches/2000-09/msg00895.html Mar 21 12:39:32 yes there is Mar 21 12:39:39 it tells GCC to save 0-7 if it uses them Mar 21 12:41:13 haha funny now arm cpus goes in to desktop gpu Mar 21 12:41:54 then the desktop GPU can have Java acceleration Mar 21 12:41:57 awesome Mar 21 12:42:11 aeh? Mar 21 12:42:42 http://en.wikipedia.org/wiki/Jazelle Mar 21 12:42:57 thumbee Mar 21 12:43:06 jazelle is dead Mar 21 12:43:27 it was my understanding that java is a banned word here Mar 21 12:43:33 maybe they'll pull an r-pi and put an arm11 in there Mar 21 12:43:41 gqq? Mar 21 12:43:53 I am the meta-java guy Mar 21 12:44:05 is that like fake coffee? Mar 21 12:44:33 'meta' sounds greek Mar 21 12:45:01 rather abstract coffee Mar 21 12:58:15 * KotH goes for the non-abstract-tea Mar 21 12:58:33 +1 Mar 21 13:11:21 my new browser homepage: http://time.is/ Mar 21 13:12:24 I should use my rpi for that Mar 21 13:12:31 and a spare 50" screen Mar 21 13:12:57 why? you like arial? Mar 21 13:15:40 I've seen several videos on using beagle boards to create video walls, but hardly any links to documentation.. can anyone suggest some resources? Mar 21 13:18:01 simple: you use the video application some dude from southampton has written, plug everthing together, sacrifice a virgin rpi user at midnight during a full moon night and it will magically work Mar 21 13:18:33 magic being the operative ingredient Mar 21 13:21:19 I get the impression you guys dislike beagleboard-based videowalls Mar 21 13:21:50 nope Mar 21 13:21:54 we dont dislike them Mar 21 13:22:00 we just like to getting paid doing them ;) Mar 21 13:22:18 ah Mar 21 13:24:34 other than that, i assume there are a metric ton of tutorials and howtos for that kind of thing Mar 21 13:24:56 it isnt like nobody has done that before Mar 21 13:27:18 foozo: a videowal with preprocessed segments can be done easily Mar 21 13:27:21 +l Mar 21 13:29:49 koth: i hardly doubt we're in the same neighbourhood ;-) care to to be a little more specific? Mar 21 13:31:46 foozo: the wall we did was using http://git.mansr.com/?p=omapfbplay;a=summary Mar 21 13:31:53 with the netsync feature Mar 21 13:32:19 seems like everyone is using omapfbplay Mar 21 13:32:27 there you go, thanks av500 :-) Mar 21 13:32:29 everyone? Mar 21 13:32:31 I think gstreamer has net-sync nowadays too Mar 21 13:32:47 yeah Mar 21 13:32:48 av500: everyone we care about Mar 21 13:33:14 av500: "everyone" not being qualified. omapfbplay showed up in several google queries. Mar 21 13:33:54 well, the TI beagle videowall is based on our videowall Mar 21 13:33:58 and that used omapfbplay Mar 21 13:34:42 didn't they have a panda-wall too? Mar 21 13:34:57 also omapfbplay Mar 21 13:35:06 sure Mar 21 13:35:26 anything else would have been surprising Mar 21 13:37:39 I wonder if the pandawall used HW decoding Mar 21 13:38:18 we could do a 8x wall on the 6678, too bad it does not have video out :) Mar 21 13:38:26 hehe Mar 21 13:38:31 you can bitbang it Mar 21 13:38:43 well, I have to see if the PCI it has is master capable Mar 21 13:38:51 then once could connect a gfx card... Mar 21 13:40:31 would surprise me if it isnt master capable Mar 21 13:40:59 TI is good for surprises Mar 21 13:41:50 juup Mar 21 13:41:53 now, how do we get nvidia to compile drivers for c6x? Mar 21 13:41:55 is there any way to multiplex the HDMI ports, so you don't need one beagleboard per display? Mar 21 13:41:55 :) Mar 21 13:42:03 biot: nope Mar 21 13:42:11 * KotH is currently rewriting all the code he has written yesterday, due such a surprise Mar 21 13:42:11 on panda you have 2 Mar 21 13:42:20 not much market for HDMI multiplexing chips, I suppose :) Mar 21 13:42:35 biot: the omap3 cannot output that many pixels Mar 21 13:42:40 av500: on the panda, you have a dvi and a hdmi... both on hdmi connectors Mar 21 13:43:35 u dont say Mar 21 13:43:36 :) Mar 21 13:46:54 * KotH knows his hardware ;) Mar 21 13:56:22 XorA: ekdbus :) Mar 21 13:56:59 :-D Mar 21 13:57:08 also uckdbus Mar 21 13:57:32 I assumed ubuntu would also want to get in on the forking game Mar 21 13:57:36 right Mar 21 13:57:43 kMirbus Mar 21 13:57:47 and I added Redhat just to complete :-D Mar 21 14:20:33 hello everyone Mar 21 14:20:50 is this the right place for uboot/mlo related questions? Mar 21 14:21:18 on beagleboards, sure Mar 21 14:21:23 ok Mar 21 14:21:38 for the community support u-boot stuff...not the TI s/w Mar 21 14:21:39 I'm working with a beagleboard xm and the TI sdk Mar 21 14:21:49 oh Mar 21 14:22:03 e2e has TI SDK knowledgeable support folks Mar 21 14:22:27 ok, in any case I'll ask in general terms ;) Mar 21 14:22:40 that works too..many issues are common Mar 21 14:22:55 so I'm trying to boot from the serial channel Mar 21 14:23:14 uploading first MLO Mar 21 14:23:18 and then uboot Mar 21 14:23:49 people still follow that cargo cult? Mar 21 14:24:17 cargo cult? Mar 21 14:24:40 http://en.wikipedia.org/wiki/Cargo_cult Mar 21 14:25:08 oh, you mean that serial communication is archaic Mar 21 14:25:18 absolutely not Mar 21 14:25:31 emilio: looking at the upstream omap3_beagle.h, it appears that SPL ymodem boot is not enabled so I'm guessing nobody is supporting that boot mode on 3530/xM beagle Mar 21 14:25:35 * ka6sox-choochoo watches...and learns Mar 21 14:26:13 yeah, I was suspecting that Mar 21 14:26:15 some of us use serial rom boot on the newer rom versions and have it enabled by default... e.g. am33xx bone, ti814x evm Mar 21 14:26:58 emilio: is it even an allowable bootmode on xM? Mar 21 14:26:58 do I have any chances enabling SPL boot myself? Mar 21 14:27:05 yes Mar 21 14:27:28 there is some documentation regarding other boards and CPUs that applies Mar 21 14:27:48 right, but the selection of boot modes hunt order is board-specific Mar 21 14:27:51 pressing the "user" button during reset allows for UART download Mar 21 14:28:08 and I follow the described protocol using omap-u-boot-utils Mar 21 14:28:34 waiting for the ASIC ID Mar 21 14:28:37 ahh, ok, I wasn't sure what all was on that boot selection Mar 21 14:28:41 then sending the MLO boot code Mar 21 14:28:57 cool, yes, google for the omap3 serial boot tools..there's like 9 different tools various people have written Mar 21 14:29:01 then sending uboot using kermit Mar 21 14:29:09 read their documentation as to which format image to squirt in Mar 21 14:29:14 yeah, you're right Mar 21 14:29:23 then look at the SPL options that the am335x_evm.h config has on for ymodem spl Mar 21 14:29:24 in my experience only omap-u-boot-utils works decently Mar 21 14:29:37 although the procedure is not very hard to reproduce by hand ;) Mar 21 14:29:38 you'll squirt the SPL image via the proprietary protocol Mar 21 14:29:45 then ymodem from spl to get u-boot.img in Mar 21 14:29:49 sure Mar 21 14:30:00 hi! Mar 21 14:30:02 assuming you have the right options on in the config file Mar 21 14:30:20 ok Mar 21 14:30:29 #define CONFIG_SPL_YMODEM_SUPPORT Mar 21 14:30:30 is it possible to compile MLO using some sort of debug mode Mar 21 14:30:37 ? Mar 21 14:30:51 #define DEBUG in the config file Mar 21 14:30:52 so that I can at least see on the serial port what is going right or wrong? Mar 21 14:31:00 will fill your log Mar 21 14:31:05 yup, that makes total sense Mar 21 14:31:08 :D Mar 21 14:31:35 ok, I'm off to try this stuff Mar 21 14:31:47 thanks mdp for your pointers Mar 21 14:31:49 good luck...several people have been successful in the past Mar 21 14:31:50 np Mar 21 14:33:18 mru, you're scaring people away from e2e ;) Mar 21 14:33:29 me? how? Mar 21 14:33:35 I've never set foot there Mar 21 14:33:44 doubt I ever will Mar 21 14:34:34 assuming that was the target of the not-too-far-off-the-mark cargo cult comment Mar 21 14:34:57 oh that Mar 21 14:35:26 I got the feeling of exact steps being followed Mar 21 14:35:53 isn't it refreshing when you aren't asked for exact steps? Mar 21 14:35:56 almost surprising Mar 21 14:36:24 is any kernel option needed to enable external usb hubs for the beaglebone? i just built linux 3.8.4-00402 (from https://github.com/beagleboard/kernel/tree/3.8), if i connect a device directly to the usb-a-port, it is detected (lsusb), but not with a hub. the hub is working fine on another machine Mar 21 14:37:54 I've been supporting other users on embedded-linux related stuff for some time Mar 21 14:38:07 I know what people want to hear when asked for help Mar 21 14:38:18 +1 Mar 21 14:38:54 just don't do this: http://thedailywtf.com/Articles/A-Misleading-Memory.aspx Mar 21 14:41:27 :D Mar 21 14:45:00 hehe Mar 21 14:45:21 funny Mar 21 14:45:33 BOOT_DEVICE_UART Mar 21 14:45:41 is only defined for CONFIG_AM33XX Mar 21 14:45:43 all that reminded me of some bad times with watcom c++ and the phar lap extender ;) Mar 21 14:46:32 whereas the beagle is OMAP34XX Mar 21 14:46:54 yeah, you'll have to do some work Mar 21 14:47:24 the first time uart boot was implemented it was done for am33xx/ti814x Mar 21 14:47:53 surely a little copying and pasting is all it takes Mar 21 14:48:01 and some ifdefs Mar 21 14:48:17 but be sure to break it for am33x Mar 21 14:48:17 yup, everything seems to be there... Mar 21 14:48:31 it'll never hit the vendor tree otherwise Mar 21 14:48:33 don't listen to mru ;) Mar 21 14:48:37 hehe Mar 21 14:48:48 yeah, should be easy to adapt Mar 21 14:49:06 check the TRM rom boot section for what value it drops for uart boot Mar 21 14:50:51 oops Mar 21 14:50:53 arm-arago-linux-gnueabi-ld: u-boot-spl section `.data' will not fit in region `.sram' arm-arago-linux-gnueabi-ld: region `.sram' overflowed by 732 bytes Mar 21 14:51:29 damn Mar 21 14:51:55 hijack the dsp l2 ram :) Mar 21 14:51:59 it's bigger Mar 21 14:52:32 could you give me exact steps on that? ;) Mar 21 14:54:33 I was also thinking about removint other features Mar 21 14:54:37 such as the nand boot Mar 21 14:55:56 mdp, is there a reason why https://github.com/beagleboard/kernel/blob/3.8/patches/audio/0003-Setup-am335x-bonelt-audio-DT.patch is set for AIC and not HDMI? Mar 21 14:57:19 joel grabbed a hebbar patch and committed that..that is just stuff for the EVM Mar 21 14:57:31 and I suspect crude hacks were made on top of that Mar 21 14:58:10 that's some pretty bizarre and misleading dts data Mar 21 14:59:35 in hebbar's tree, he was trying to target am335x-evm that has the AIC and use the legacy davinci machine file...adding DT parsing support to that. Mar 21 15:04:14 mdp: btw, c6x code from the ti compiler is _much_ faster than what gcc produces Mar 21 15:04:30 the bits I've been able to run, that is Mar 21 15:04:44 you should post a performance comparison on e2e Mar 21 15:04:52 * mdp ducks Mar 21 15:05:19 the fft in libav runs >2x faster with the ti compiler Mar 21 15:05:26 and the self-tests pass Mar 21 15:05:38 so either it's working, or it's failing very cleverly Mar 21 15:05:57 :) Mar 21 15:06:20 mru: but gcc being rather slow was expected, no? Mar 21 15:06:24 of course Mar 21 15:06:41 both float and fixed-point btw Mar 21 15:10:59 Hi! Mar 21 15:11:14 mdp, I have revamped it but didn't want to commit it if there was a reason to keep that. Mar 21 15:11:28 I have revamped it but didn't want to commit it if there was a reason to keep that. Mar 21 15:11:32 whoops Mar 21 15:11:44 I was wondering if a beagleboardxm would be suitable for building a desktop pc Mar 21 15:11:58 no Mar 21 15:11:59 the 1ghz of cpu is enough for me as I use a pentium III 750mhz Mar 21 15:12:08 also https://github.com/beagleboard/kernel/blob/3.8/patches/audio/0006-ASoC-Davinci-machine-Add-device-tree-binding.patch is totally misleading...I've revamped it to include am335x evm but I am thinking this is the old way and I should commit a DT version that I've been playing with. Mar 21 15:12:25 ka6sox-choochoo: only thing I wonder is if your changes are compatible with panto's commited version of the dvi+audio cape support Mar 21 15:12:38 but is there any other feature of the bleagle that could make it unsuitable for desktop use? Mar 21 15:12:48 slow CPU Mar 21 15:12:51 little memory Mar 21 15:12:52 usb Mar 21 15:12:55 slow mass storage Mar 21 15:13:02 it has less than 8 cores and less than 16GB so it can't run Unity Mar 21 15:13:07 apart from that its fine Mar 21 15:13:08 I guess what I'll do is create separate HDMI versions. Mar 21 15:13:12 I am using a pentium 3 750mhz 500mb ram Mar 21 15:13:26 what year is it? Mar 21 15:13:41 mdp, the modes are not compatible. Mar 21 15:13:41 and I'm not sure which one is "default" Mar 21 15:13:43 * mdp fires up the blackhawk 603e Mar 21 15:13:46 i use a very plain debian with ratpoison and stuff like that Mar 21 15:14:25 aaragui: well, if you insist you can of course use it Mar 21 15:14:56 ka6sox-choochoo: well, I guess hdmi audio is the default on black.. Mar 21 15:15:00 when you say that the mass storage is slow, you mean slow as in painfully slow? Mar 21 15:15:08 slow as usb slow Mar 21 15:15:17 I would think that "onboard" would be default and a cape would apply only if the cape is present. Mar 21 15:15:18 and an inserted dvi+audio cape should override that and use mcasp with the aic on the cape Mar 21 15:15:23 right Mar 21 15:15:27 but since you appear to have already made up your mind, just go ahead Mar 21 15:15:32 we are in violent agreement Mar 21 15:15:43 could it play a movie in vlc for example? (sorry for asking so much) Mar 21 15:16:25 dvd-quality should work ok Mar 21 15:16:25 mdp then I'd like to modify the default dts to include the pinmuxing for HDMI audio and not the SPI currently there. Mar 21 15:16:32 720p if you're lucky Mar 21 15:16:47 ka6sox-choochoo: you might need to chat with koen about that Mar 21 15:17:01 thank you very much guys Mar 21 15:17:04 I will get one Mar 21 15:17:04 mdp, I will Mar 21 15:17:07 I seem to remember the spi muxing and spidev being in there to help with a user experience thing Mar 21 15:17:17 brb Mar 21 15:17:23 but I would think that should move to an overlay if it hasn't already Mar 21 15:17:47 koen, ping? Mar 21 15:17:58 seems like a trivial little overlay to document loading for people wanting to rock out with some spidev play Mar 21 15:18:21 ka6sox-choochoo: ping Mar 21 15:19:47 koen, I'd like to modify the arch/arm/boot/dts/am335x-boneblack.dts to use HDMI pinmixing default (as its onboard) and let the dvi+audio cape mod that if present. Mar 21 15:20:41 koen, tangential note: I'm so pleased the my commit to fix the dts name has resulted in the name being repeated by countless people by reference ;) Mar 21 15:20:51 multiplier effect ftw Mar 21 15:22:14 I'm about a minute from going underground...cu on the other side Mar 21 15:22:35 znc will pick up any answers Mar 21 15:22:39 heh Mar 21 15:22:48 hiya Tartarus Mar 21 15:23:07 hey ka6sox-choochoo Mar 21 15:23:59 glad I took the train today...getting a lot of work done on this audio stuff. Mar 21 15:25:37 * ka6sox-choochoo awaits word from on high :) Mar 21 15:27:22 ka6sox-choochoo: we need it as a cape to support LCDs as well Mar 21 15:27:42 * emeb thought that said "ka6sox-achoo" and almost said Mar 21 15:27:48 bless you Mar 21 15:28:20 ka6sox-choochoo: hope you don't get a cinder in your eye...keep that window up! Mar 21 15:31:27 koen the default pins are for SPI audio unless thats SPI LCD instead Mar 21 15:31:27 mdp, window is UP Mar 21 15:31:39 ;) Mar 21 15:31:53 ka6sox-choochoo: throw some more coal in the burner please Mar 21 15:32:10 koen, if LCD is present wouldn't capemanager reconfigure? Mar 21 15:32:11 I've always been taught that you default configure for what is onboard..any externally hooked up thing is managed later. Mar 21 15:34:47 am I here? Mar 21 15:35:27 no Mar 21 15:35:45 :) Mar 21 15:36:21 ka6sox-choochoo: we don't consider hmdi and emmc onboard :) Mar 21 15:36:27 they are virtual capes :) Mar 21 15:37:00 oh Mar 21 15:37:07 hmmm.... Mar 21 15:37:07 that confuses things further... Mar 21 15:37:17 so in the absense of any *real* capes you default to those or what? Mar 21 15:37:19 koen: what about GPTIMER12? Mar 21 15:37:23 is that virtual too? Mar 21 15:38:39 koen, does that mean you will support them with virtio and virtdma? Mar 21 15:39:33 koen, where is the configuration eeprom for the "virtual" capes? Mar 21 15:47:15 virtual inside capemgr Mar 21 15:48:09 in the absence of any external cape why not just default to the onboard? Mar 21 15:55:18 ka6sox-choochoo: that's what it's currently doing Mar 21 15:55:55 mdp: you mentioned before something about hijacking dsp memory Mar 21 15:56:18 i'm really new to the beaglebone, but i'm trying to attach an atmel at86rf230 radio to the beaglebone. i've got some sample code from the raspberry pi. i won't need a cape with eeprom, just for testing purposes. in which file do i have to set the pins? is arch/arm/mach-omap2/board-omap3beagle.c the one i'm looking for or is it just for the beagleboards, not the bone? Mar 21 15:56:24 could you explain that a bit further? Mar 21 15:56:31 for boneblack.dts it is configured for SPI which will confuse the connected device Mar 21 15:56:45 it should be defaulted to i2s as thats what is hooked up Mar 21 16:01:38 hmm, now running ./avconv gives me "incorrect codes" Mar 21 16:02:25 can someone give me the correct codes? Mar 21 16:02:57 emilio: sure it was me? I don't recall atm Mar 21 16:03:29 hijack the dsp l2 ram :) Mar 21 16:03:40 sorry, no, it was mru Mar 21 16:03:52 why do you guys have such confusing names? Mar 21 16:05:46 somebody already had "Kraken" so I had to stick with initials Mar 21 16:06:40 yeah... those damn irc name squatters Mar 21 16:09:11 okay peeps and trolls..I'm off to work for the day... Mar 21 16:09:34 emilio: why again do you need more ram? Mar 21 16:09:44 xloader over serial should work Mar 21 16:10:00 av500: not anymore, really Mar 21 16:10:12 I was trying to #define DEBUG Mar 21 16:10:18 and? Mar 21 16:10:23 for the omap3_beagle Mar 21 16:10:40 and MLO was getting too large Mar 21 16:11:01 but I've disable some other stuff and now it fits Mar 21 16:11:01 then drop some stuff Mar 21 16:11:07 right Mar 21 16:11:17 if you boot from serial drop the MMC read stuff Mar 21 16:11:18 but I still see no debug stuff on the serial port Mar 21 16:11:19 and FAT Mar 21 16:11:32 do you see non-debug stuff on the serial? Mar 21 16:11:44 not really Mar 21 16:11:53 I'm sending MLO Mar 21 16:12:11 which serial port is used for debug data? Mar 21 16:12:24 UART3? Mar 21 16:12:29 tty=2 Mar 21 16:12:32 ttyO2 Mar 21 16:12:47 with fortran counting that might be 3 Mar 21 16:12:47 hmmm, that's the one I'm using to download the MLO image Mar 21 16:13:03 so? Mar 21 16:13:31 I download it and I fire up minicom on that same port immediately after Mar 21 16:13:35 but nothing shows up Mar 21 16:14:11 you can start your serial downloader from within minicom Mar 21 16:14:50 that vaguely sounds familiar... Mar 21 16:14:59 CTRL-A S Mar 21 16:14:59 run script (Go)? Mar 21 16:15:14 that's for sending files Mar 21 16:15:19 yes Mar 21 16:15:27 but you can define the program to send Mar 21 16:15:37 I used that with my loaders for years Mar 21 16:16:08 I only get a predefined set of protocols: xmodem, kermit, etc. Mar 21 16:16:31 you can edit Mar 21 16:16:50 CTRL-A O Mar 21 16:16:54 file transfer protocols Mar 21 16:16:56 oh right Mar 21 16:17:08 set filename to N Mar 21 16:17:19 add your script with fixed filename Mar 21 16:17:41 and I/O redirection to Y Mar 21 16:17:55 then you loader talks to stdin/out only Mar 21 16:18:07 and minicom pipes that to the serial port Mar 21 16:18:26 I see Mar 21 16:18:38 of course it might be an issue if the baud is different Mar 21 16:18:51 not the baud Mar 21 16:18:55 parity is different Mar 21 16:19:18 Tartarus and I just use screen with the ctrl-a : escape to run the serial boot script Mar 21 16:19:33 yeah Mar 21 16:19:38 j Mar 21 16:19:46 $ cat ~/bin/load-boot-xymodem.sh Mar 21 16:19:46 #!/bin/sh Mar 21 16:19:46 /usr/bin/sx -kb8 /tmp/u-boot-spl.bin Mar 21 16:19:46 /usr/bin/sx -kb8 --ymodem /tmp/u-boot.img Mar 21 16:19:47 or that Mar 21 16:20:02 and I copy spl/u-boot-spl.bin and u-boot.img to /tmp as part of my build line Mar 21 16:20:06 modified for your non xmodem method of rom boot on dm3730, of course Mar 21 16:20:59 ok Mar 21 16:21:15 god i hat this ti sdk so much Mar 21 16:21:48 everything is packed for just one single purpose Mar 21 16:22:14 you have other ideas in mind? well, bad luck, everything is broken Mar 21 16:22:39 emilio: then you can hire a lot of ex-TI people to help you Mar 21 16:23:17 hungry people. Mar 21 16:23:33 emilio: it could be worse...it could be an android sdk Mar 21 16:23:38 we did that for production flashing Mar 21 16:23:46 via USB Mar 21 16:24:17 * av500 wonders where ssc is these days Mar 21 16:24:49 yeah, sad affair, the layoffs Mar 21 16:25:20 so, there's something I'm a bit confused about Mar 21 16:25:28 when booting via UART Mar 21 16:25:39 which file should be uploaded first? Mar 21 16:25:41 its ok to be confused Mar 21 16:25:46 MLO Mar 21 16:25:50 MLO o u-boot-spl.bin? Mar 21 16:25:54 or whatever you call your first stage Mar 21 16:25:57 same thing Mar 21 16:26:05 just header difference Mar 21 16:26:08 in fact, the filename itself is not uploaded Mar 21 16:26:29 on the am335x/omap4/ti814x rom we load u-boot-spl.bin...it reads a raw binary Mar 21 16:26:43 but uboot build creates both Mar 21 16:27:05 it's being nice..giving you an MLO for mmc purposes, for example Mar 21 16:27:07 an MLO and then a u-boot-spl Mar 21 16:27:12 the .bin is an intermediate image Mar 21 16:27:27 u-boot-spl is the elf file Mar 21 16:27:48 so I have transfer u-boot-spl.bin raw, and the u-boot.img via ymodem? Mar 21 16:28:28 emilio: I have no idea what your DM3730 rom wants for uart boot Mar 21 16:28:33 it's dependent on your ROM requirements Mar 21 16:28:54 it's completely different from the new ROMs Mar 21 16:28:57 I've checked it against the manual Mar 21 16:29:01 and? Mar 21 16:29:09 and only the MLO file matches what seems to be required Mar 21 16:29:21 ok, so it requires some header Mar 21 16:29:40 doublecheck it, it's critical Mar 21 16:29:48 yes Mar 21 16:29:54 so once the MLO file is there Mar 21 16:29:56 what is next? Mar 21 16:29:59 plus, docs on several of the cli tools created for loading images should also explain what image to load Mar 21 16:30:02 u-boot.img? Mar 21 16:30:07 u-boot.img, yes Mar 21 16:30:16 just like the script Tartarus pasted Mar 21 16:30:18 ok Mar 21 16:30:23 the second line will work Mar 21 16:30:39 the first you need to change to load using your dm3730 method Mar 21 16:31:17 Yeah, I've not tried UART loading on dm3730, but in theory it should be easily supported on mainline u-boot for SPL Mar 21 16:31:23 might need to populate a few defines Mar 21 16:31:28 what about uboot-spl.bin? Mar 21 16:31:33 not necessary? Mar 21 16:31:45 emilio: On am335x/etc, the UART mode specifies a raw file, no header Mar 21 16:31:57 emilio: uboot spl is a first stage Mar 21 16:31:58 If dm3730 says it wants the same header as MMC, then MLO is your file, ignore spl/u-boot-spl.bin Mar 21 16:32:02 much like MLO Mar 21 16:32:09 u-boot SPL replaced x-loader Mar 21 16:32:24 Both provide "MLO" which is a first stage loader with a header on top Mar 21 16:32:26 ok, that makes sense Mar 21 16:32:29 and "MLO" as a name is only needed when loaded from FAT Mar 21 16:33:22 yes, i know Mar 21 16:33:32 * mdp pops out for errands Mar 21 16:34:04 so, what im doing is: Mar 21 16:34:05 ./pserial -p /dev/ttyUSB0 -f ../u-boot-2012.04.01/build/MLO Mar 21 16:37:55 wee, falcon boot on the bone Mar 21 16:38:45 that wouldbe awesome on the hawkboard Mar 21 16:39:05 eagleboard! Mar 21 16:39:29 eagle eats falcon for breakfast Mar 21 16:39:35 Tartarus: strangely enough it doesn't having the uImage loaded over the bss/malloc section ;) Mar 21 16:40:13 details Mar 21 16:44:32 mru: yeah Mar 21 16:45:24 but ok, atleast it works: http://pastebin.com/wmDKakwF Mar 21 16:46:17 Jacmet: why would that not work? Mar 21 16:46:50 but - does it load uEnv.txt? :) Mar 21 16:46:57 bradfa: public shaming is only reason to get better :) Mar 21 16:46:59 ok guys, I'm off for the day Mar 21 16:47:06 thanks a lot for the help Mar 21 16:47:27 mranostay, I wouldn't say it's the _ONLY_ reason to get better... Mar 21 16:47:28 :) Mar 21 16:47:34 av500: it loads 'args' ;) A binary dump of the atags/dtb Mar 21 16:48:04 someone is asking for a a public shaming Mar 21 16:48:23 bradfa: which for me is everytime i post gerbers :) Mar 21 16:48:26 * av500 shames mranostay in public Mar 21 16:48:41 * bradfa throws stones at mranostay in the town square! Mar 21 16:49:03 did he say y****a? ;) Mar 21 16:49:12 av500: a bit fiddly as the u-boot fat write support seems a bit flaky, but it's short enough to hexdump/recreate on a pc Mar 21 16:49:28 ? Mar 21 16:49:31 FAT write? Mar 21 16:49:43 Jacmet: nice Mar 21 16:49:45 av500: yes, for spl export Mar 21 16:49:54 * av500 is confused Mar 21 16:49:59 that's been on my todo list for a while Mar 21 16:50:17 so uboot writes a SPL copy of itself to the sd card? Mar 21 16:50:18 You should be able to get it to read uImage from ext* with a little work :) Mar 21 16:50:26 av500: No, so, "Falcon mode" Mar 21 16:50:29 It's SPL boots Linux Mar 21 16:50:31 Tartarus: I've moved bss/spl malloc area past the 8MB mark instead Mar 21 16:50:40 and part of that is doing some prep work in U-Boot Mar 21 16:50:43 Tartarus: yes, I get that Mar 21 16:50:47 so that we save the junk we pass to the kernel Mar 21 16:50:48 ATAGS Mar 21 16:50:50 or DT Mar 21 16:50:51 and such Mar 21 16:50:53 av500: spl export does a dump of the atags/adjusted dtb Mar 21 16:50:54 in the ready to go form Mar 21 16:51:11 spl is the u-boot command for that Mar 21 16:51:15 a tad confusing, yes Mar 21 16:51:16 ah Mar 21 16:51:20 well Mar 21 16:51:22 Tartarus: any specific reason you put it bss/malloc at bottom of ram? Mar 21 16:51:39 Tartarus: so I need to normal boot once before I can falcon? Mar 21 16:51:47 av500: yes Mar 21 16:51:53 Jacmet: Not much, no, So long as you shuffle things so it still works on 256MB platforms, go ahead Mar 21 16:51:54 to feed the falcon :) Mar 21 16:52:10 av500: the idea of falcon boot is to have a normal dev mode and fast production mode Mar 21 16:52:49 Tartarus: I just offset it 10MB, so that should be fine Mar 21 16:53:41 av500: on the bone, I simply check for a 'c' character on uart0 at boot, and go to full u-boot mode if detected, otherwise direct linux boot Mar 21 16:54:05 right Mar 21 16:54:10 or a GPIO on prod hw or so Mar 21 16:54:26 av500: yes Mar 21 16:54:35 Jacmet: I'm more used to hard coding stuff like tags and DTBs inside kernesl Mar 21 16:54:57 I know that is not what lkml things is the right way Mar 21 16:55:01 thinks Mar 21 16:55:27 av500: the nice thing imho about falcon mode is that it is so easy to switch between development and product mode Mar 21 16:55:42 sure Mar 21 16:55:42 koen: poke poke Mar 21 16:55:51 * mranostay pokes av500 accidently Mar 21 16:56:07 as long as its not a stab Mar 21 16:56:08 ... Mar 21 16:56:30 too late Mar 21 16:56:52 Jacmet: thanks for doing it, this will be a nice thing on the voldemort board for example :) Mar 21 16:56:59 koen: any time for a few more updates? :) Mar 21 16:57:02 Tartarus: voldemort? Mar 21 16:57:09 Jacmet: inside joke Mar 21 16:57:23 the dark bone Mar 21 16:57:28 Right Mar 21 16:57:33 since we aren't allowed to mention the color Mar 21 16:57:34 The next gen beagleboard we can't name Mar 21 16:57:42 But there's various things known / public Mar 21 16:57:44 it's BLACK Mar 21 16:57:46 we can say it's not white Mar 21 16:57:50 including hey, pretty full u-boot support Mar 21 16:57:51 nor mauve Mar 21 16:58:18 Tartarus: are you happy with the spl_mmc type cleanup? If so, then I'll base my spl-mmc-falcon stuff on it Mar 21 16:58:26 it is burnt Mar 21 16:58:34 av500: ahh, that one ;) Mar 21 16:58:49 Jacmet: I think so, lemme reviewed-by it quick Mar 21 16:58:52 ds2: it passed a thorough burnin test Mar 21 16:58:59 it's BLACK Mar 21 16:59:01 every bit of the 2GB NAND tested Mar 21 16:59:03 you can get the new board by plugging in the phoenix cape to apply 25KV to the current board Mar 21 16:59:10 :) Mar 21 16:59:26 ds2: students can use just 12V Mar 21 17:00:23 only problem with falcon boot is if you use external dtb, that it more or less gets baked into the board Mar 21 17:00:33 and we all know how stable dtb is Mar 21 17:00:50 av500: the phoenix process requires a higher activation energy Mar 21 17:00:59 Jacmet: exactly Mar 21 17:01:05 hence the in-kernel DTB Mar 21 17:01:10 for prod stuff Mar 21 17:01:14 F the DT Mar 21 17:01:17 just patch it out Mar 21 17:01:42 * av500 meets ds2 in the dark to hand over an illegal board file Mar 21 17:02:24 so that's who's in front of those bright headlights! Mar 21 17:02:46 put the board file down and walk away slowly Mar 21 17:03:00 patches up where I can see them Mar 21 17:03:52 ds2: heretic! Mar 21 17:04:06 hereDTic Mar 21 17:04:21 ds2: the real fun will start with chinese DT Files Mar 21 17:04:30 or DT support in general Mar 21 17:05:11 I expect a one liner saying: compatible: "random tablet name" Mar 21 17:06:06 av500: you kid... Mar 21 17:06:17 av500: or unicode characters :) Mar 21 17:06:42 well, I guess DT supports that Mar 21 17:06:45 no? Mar 21 17:07:15 haven't been insane or drunk enough to try Mar 21 17:07:26 I also expect one DT file in the bootlaoder, one in the kernel and neither one used Mar 21 17:07:29 mranostay: that should be easily fixed Mar 21 17:07:42 mru: you make him drunk, I make him ... Mar 21 17:07:59 er what is ... Mar 21 17:08:03 mranostay: coming to elce? Mar 21 17:08:37 mru: in a suit I heard Mar 21 17:09:01 then he's already insane Mar 21 17:09:06 our job is half-done Mar 21 17:10:15 what is the other half? Mar 21 17:10:30 getting you drunk Mar 21 17:11:12 never been drunk in a suit that iirc Mar 21 17:11:22 there's a first time for everything Mar 21 17:11:45 chances are you won't remember it of course Mar 21 17:13:07 i'm sure you guys will fill me in on the interesting parts Mar 21 17:25:02 hmm, could we start friday a bit early? Mar 21 17:28:15 +1 Mar 21 17:28:17 +20 Mar 21 17:28:28 no +infinity Mar 21 17:28:33 and what about skipping Friday altogether? Mar 21 17:28:35 what? the dreadul weekend starting early? Mar 21 17:28:39 dreadful Mar 21 17:28:49 indeed Mar 21 17:29:03 who'd want that? Mar 21 17:29:10 <- Mar 21 17:29:10 no signs of anything interesting happening this weekend Mar 21 17:29:18 somebody working in a salt mine maybe Mar 21 17:29:26 but then, they dont even have weekends Mar 21 17:29:27 you rang? Mar 21 17:29:36 mranostay: go back and dig! Mar 21 17:30:02 * mranostay still can't believe the CEO used that in his all-hands Mar 21 17:31:13 hmm? Mar 21 17:31:28 mru: he used the term salt mines in his all-hands meeting Mar 21 17:31:34 who? Mar 21 17:31:44 CEO of the company i work for Mar 21 17:32:18 in what context? Mar 21 17:32:53 "i like salts mines i think they are productive" swear to god :) Mar 21 17:33:23 throw him into one Mar 21 17:36:39 mranostay: time to move on w/CEO like that Mar 21 17:37:10 kkeller: no @#@$# :) Mar 21 18:43:43 _av500_: as were you part of that 36k check i saw yesterday? :) Mar 21 20:38:16 <_av500_> mranostay: ? Mar 21 20:39:18 av500 Mar 21 20:39:19 ? Mar 21 20:39:24 _av500_ Mar 21 20:39:49 Shadyman: ! Mar 21 20:39:57 <_av500_> ¿ Mar 21 20:41:38 mru Mar 21 20:59:34 thats it? Mar 21 21:03:01 ? Mar 21 21:06:22 …………! Mar 21 21:07:39 ¿?¿?¿?¿?¿?¿? Mar 21 21:20:43 Has development of kernel support for the BBxM in the oe/angstrom toolchain stalled? Mar 21 21:21:27 what has the toolchain todo with kernel programming? Mar 21 21:21:43 can't do one without the other Mar 21 21:22:13 the kernel is done Mar 21 21:22:14 I'm talking about the recipes that make building kernels relatively simple for some of us Mar 21 21:22:22 that kernel doesn't compile itself, you know Mar 21 21:22:35 make menuconfig; make Mar 21 21:22:40 how could it possibly be simpler? Mar 21 21:22:51 Ok, to be clear, virtual/kernel hasn't budged from 3.2.28 in over a year Mar 21 21:23:01 smart relax doesn't work Mar 21 21:23:05 reflex Mar 21 21:23:44 etc, As I've gathered all of these things are things that have been fixed in patches since then, I've certainly seen them discussed. Mar 21 21:24:49 oh, they've been discussed alright Mar 21 21:24:56 doesn't mean they work Mar 21 21:24:58 I got all excited about the recipes for mainline linux 3.7 & 3.8 that are now in meta-beagleboard, but then noticed just now that they are specifying machine beaglebone Mar 21 21:25:14 if you want something that _works_, you'll have to go for the 2.6.22 vendor kernel Mar 21 21:25:24 or was it .18? Mar 21 21:25:31 <_av500_> I thought 3.8 work on all beagles Mar 21 21:25:41 I thought so to Mar 21 21:25:42 for moderate values of work Mar 21 21:25:47 do they have full pm? Mar 21 21:26:26 ah nice phone starts up with 4.2.2 Mar 21 21:26:30 yea, but going back to 2.6.22 loses all the v4l kernel work with omap3isp Mar 21 21:27:01 oh my, I seem to have left off a smiley! Mar 21 21:27:11 how will this end? Mar 21 21:28:23 anyway, I guess I've just gotten a little confused about angstrom 2012.12 meta-beagleboard, and what improvements have actually been made Mar 21 21:29:42 I don't think I actually need 3.8, I thought that the pm patches and smart reflex stuff had made it into mainline Mar 21 21:30:07 but if they haven't, then I guess there is no point screwing with it. Mar 21 21:34:11 has anyone taken the time to move to a version of connman more recent than 0.79? Mar 21 21:36:58 * Russ chuckles https://www.codeaurora.org/gitweb/quic/le/?p=kernel/msm.git;a=commitdiff;h=d21ced2f0dda81 Mar 21 21:37:25 russ let's see Mar 21 21:37:57 yes Mar 21 21:38:00 there are fucked up things, and there are fucked up things Mar 21 21:38:05 the python wrapper Mar 21 21:40:27 wait until python makes it *in* the kernel so you can more easily attach kdbus plugins ;) Mar 21 21:41:09 <_av500_> https://twitter.com/politbuntu Mar 21 21:41:48 LOL Mar 21 21:41:55 <_av500_> will the REAL_CC please shut up Mar 21 21:42:12 okay.jpg Mar 21 21:44:46 <_av500_> onlya wrapper with a good warning can stop a compiler with a bad warning Mar 21 21:47:25 <_av500_> this is cool website design: http://www.control.com/thread/1353519492 Mar 21 21:48:18 _av500_: a rapper wrapper? Mar 21 21:48:21 dawg Mar 21 21:48:35 gees, the ninties ! Mar 21 21:49:02 "advertise here" Mar 21 21:49:15 <_av500_> ogra_: unpossible, they are discussing android Mar 21 21:49:31 so back to the future then Mar 21 21:49:34 :) Mar 21 21:50:02 its a wormhole between the ninties and android :) Mar 21 21:54:25 <_av500_> damn Mar 21 21:54:33 <_av500_> the bone does not have a working FIQ Mar 21 21:54:38 <_av500_> they effed it up Mar 21 21:54:44 FIQ That. Mar 21 21:54:52 <_av500_> yeah Mar 21 21:59:44 <_av500_> interesting: http://e2e.ti.com/support/arm/sitara_arm/f/791/t/207606.aspx Mar 21 21:59:55 <_av500_> that links to public PDFs that describe MUSB Mar 21 22:01:18 <_av500_> I always thought that was s3cr3t Mar 21 22:04:11 hahahhe Mar 21 22:04:14 so fix it Mar 21 22:04:22 ok, so it looks lithe angstrom-v2012.12-yocto1.3 branch of the setup-scripts configures layers.txt for a branch that has connman 1.4 Mar 21 22:04:54 spacecolonyone and now? Mar 21 22:05:06 I don't follow Mar 21 22:05:33 can someone give a quick rundown on the branches master, next, angstrom-v2012.12-yocto1.3, & angstrom-v2012.05-yocto1.2 Mar 21 22:07:09 no Mar 21 22:07:10 One thing that I think I'm seeing that the the oebb.sh script is changing the way it handles the layers being on different branches than it expected based on layers.txt Mar 21 22:07:27 and good nite Mar 21 22:15:18 Guys, was I out of line? Mar 21 22:15:37 <_av500_> no Mar 21 22:15:45 <_av500_> but this is not #oeexperts :) Mar 21 22:16:26 eh, he seemed peeved at me in #oe as well. Mar 21 22:16:58 I asked in #angstrom, but that room is almost always dearly quiet Mar 21 22:18:12 _av500_: koen lives here, koen knows all. :D Mar 21 22:18:14 spacecolonyone_: pick a release branch of the setupscripts and follow the README there Mar 21 22:18:38 I left #angstrom when too many kiddies there tried to explain to me how angstrom actually worked Mar 21 22:18:38 speaking of... Mar 21 22:19:33 yea, I noticed the read me in some of the branches finally said do . .oe/enviroment... Mar 21 22:19:38 I do read them Mar 21 22:19:55 So then, in short angstrom-v2012.12-yocto1.3 is considered a release branch Mar 21 22:20:03 it is Mar 21 22:20:18 and master currently the same as v2012.05 Mar 21 22:20:27 good Mar 21 22:20:36 because that is what i have been using all this time Mar 21 22:20:45 but i want a less buggy connman version Mar 21 22:21:04 and it looks like 2012.12 gives that Mar 21 22:21:53 I'm not good enough to tell for sure it will be built, by inspection, but i see 1.4 is there in recipe form, so I can force it if I must Mar 21 22:24:29 koen, mru mentioned that smart reflex stuff never made it in again after the 2.6.22 PSP or whatever, I thought I'd seen otherwise somewhere along the way. Would you just say I give up hope for it unless i do it self? Mar 21 22:24:58 … or go to an old kernel Mar 21 22:25:06 I didn't say quite that Mar 21 22:25:14 my bad Mar 21 22:26:01 doesn't smart reflex require an i2c bus that isn't exposed? Mar 21 22:26:52 spacecolonyone_: 3.8 has the beginning of the ABB stuff you need for 1GHz Mar 21 22:27:18 but there doesn't seem to be a recipe in any of the branches that does 3.8 nor the BBxM Mar 21 22:27:38 and I don't think I need it badly enough to justify the time writing it. Mar 21 22:28:13 the linux-mainline 3.8 recipe for bone actually works on xM as well Mar 21 22:28:38 but I haven't activated that yet since DT overlays need to get 'ported' to regular beagle first Mar 21 22:28:48 So bitbake virtual/kernel is still building 3.2.28 Mar 21 22:28:50 ahh... Mar 21 22:28:53 DT ? Mar 21 22:28:59 devicetree Mar 21 22:29:09 cures cancer, brings world peace Mar 21 22:29:21 I guess what I'm asking boils down to: Should I just accept my 3.2.28 kernel on BBxMrevc and call it a day. I realize constantly chancing the latest "greatest" version is probably pretty Quixotic without reason Mar 21 22:29:27 harsher punishments for parole violaters, the works Mar 21 22:29:28 hah Mar 21 22:29:53 3.2 has the most "just works" bits currently Mar 21 22:29:57 does activating it consist of adding it to the MACHINE list in the recipe or something else Mar 21 22:30:30 change COMPATIBLE_MACHINE to include beagleboard Mar 21 22:30:31 I was trying to figure out why/how it was that bitbake wasn't selecting it, but haven't figured it out yet Mar 21 22:30:38 that should do the trick iirc Mar 21 22:30:43 * koen zzz now Mar 21 22:30:52 gotcha, cheers! Mar 21 22:50:20 oh, so systemd swallows dbus ? Mar 21 22:50:30 Xorg next then ? Mar 21 22:51:30 then the kernel? Mar 21 22:53:00 kdbus is in the kernel, systemd hooks into it Mar 21 22:55:00 * ogra_ points to http://lists.freedesktop.org/archives/systemd-devel/2013-March/009797.html Mar 21 22:55:54 ah Lennart again Mar 21 22:56:39 yep Mar 21 22:58:52 what is with that kid? Mar 21 23:03:00 power supply drama Mar 21 23:03:03 urgh. Mar 21 23:03:33 the two heat sinks on this device should not be shorted Mar 21 23:03:50 neither is attached to AC ground Mar 21 23:03:59 wow, thats evil Mar 21 23:04:16 is that homebrewed ? Mar 21 23:04:20 no.. Mar 21 23:04:29 beryllium oxide to the rescue Mar 21 23:04:41 its a Meanwell UL approved Power supply. Mar 21 23:04:59 mrpackethead: that sounds toxic as all hell Mar 21 23:05:22 "approved" ... Mar 21 23:05:33 Meanwells psu's are normally very good Mar 21 23:05:49 so, i'm sure i'm just missing something Mar 21 23:06:53 http://www.meanwell.com/search/PPS-200/default.htm Mar 21 23:07:08 i want to stick it in a sealed alluminum enclosure Mar 21 23:08:14 my preference woudl be to bolt the heat sinks to the case Mar 21 23:08:46 which will act as a pretty good and big heat sink Mar 21 23:15:32 that's where the beryllium oxide comes in Mar 21 23:22:08 mru: what does that smell like? Mar 21 23:22:36 the solid form, nothing I afaik Mar 21 23:30:22 if i put just mount the PSU in teh box Mar 21 23:30:27 on stand-offs Mar 21 23:30:33 and allow conveciton cooling Mar 21 23:30:38 i doubt it will work Mar 21 23:32:01 if you're afraid of beryllium oxide you can always use diamonds Mar 21 23:32:13 yeah.. Mar 21 23:32:17 i'm just afraid of the cost Mar 21 23:32:31 i'm sure that is much cheaper.. Mar 21 23:33:19 this could be a case of, opps, i've not thought this through very well Mar 21 23:34:50 mrpackethead: can i cover my r-pi in diamonds to reduce heat? Mar 21 23:35:41 er mru Mar 21 23:36:19 mranostay: are you asking if it's possible in general or if _you_ can do it? Mar 21 23:36:56 both Mar 21 23:37:33 is diamond coating armv6 compatible ? Mar 21 23:38:21 sure, if it's hardfloat Mar 21 23:38:30 * mranostay faceplams Mar 21 23:38:33 ah Mar 21 23:38:50 someone git added all the *.ko files to the repo Mar 21 23:39:14 heh Mar 21 23:39:23 saves you from building :) Mar 21 23:39:32 one step less :) Mar 21 23:39:41 maybe for people that want to cut the build time and have extra bandwidth :P Mar 21 23:39:44 i need beer.. Mar 21 23:40:09 * mru is out of beer Mar 21 23:40:43 still some scotch left in the bottle... Mar 21 23:40:57 cheers Mar 21 23:41:34 mru: pass the bottle sir Mar 21 23:43:40 ogra_: who needs shotglasses when you have a bottle? :) Mar 21 23:43:53 haha Mar 21 23:44:17 * ogra_ uses water glasses over here though Mar 21 23:46:31 russia? Mar 21 23:46:41 germany :) Mar 21 23:46:47 same difference Mar 21 23:46:52 * mranostay ducks Mar 21 23:47:32 yeah, for someone whose world ends at the shoreline for sure :P Mar 21 23:48:05 nah i'm part german so roughly know where it is on the map :P Mar 21 23:49:13 * ogra_ hands mranostay some of the laphroig Mar 21 23:50:01 +a Mar 21 23:56:12 if you do a 'static uint64_t value = 0;' inside a _function_ please die Mar 21 23:57:33 lol, in C++ or standar C ? Mar 21 23:57:38 standard c Mar 21 23:57:49 it's convenient in standard C Mar 21 23:58:18 not so much in C++ where it might as well be member variable Mar 21 23:59:17 thurbad: what is the point if you are setting to a value everytime? Mar 21 23:59:30 statics only set the first time Mar 21 23:59:55 mru: walking c99 dude Mar 22 00:00:44 the second time you came into the function it would be the last value, and would not get reinitialized Mar 22 00:01:21 it's kind of like making it a global, without the nasty scope Mar 22 00:06:39 http://www.ecnmag.com/articles/2012/12/thinking-inside-box-getting-heat-out-sealed-enclosures Mar 22 00:07:30 cute Mar 22 00:08:04 mranostay: what thurbad said Mar 22 00:09:25 I dislike that static has diffferent meanings in different contexts... but it is what it is Mar 22 00:09:42 not really Mar 22 00:09:47 static always means the same thing Mar 22 00:09:53 it's the lack of it that varies Mar 22 00:09:54 not in C++ Mar 22 00:10:05 nor in fortran, so what? Mar 22 00:10:15 static member, methods Mar 22 00:10:32 I'm talking about c, nothing else Mar 22 00:10:37 ok Mar 22 00:10:38 c++ is just fucked up Mar 22 00:10:42 lol Mar 22 00:10:55 even a comma can mean anything there Mar 22 00:10:59 yes, really Mar 22 00:11:23 C++ makes it slightly easier to scale... or chop your own head off depending on how you useit Mar 22 00:11:52 C gives you enough rope to hang yourself, C++ ties the noose for you too Mar 22 00:12:05 I don't disagree Mar 22 00:12:23 mru: what kicks the chair? Mar 22 00:12:24 but if mass hangings are your thing, sure it scales better Mar 22 00:12:42 mranostay: javascript Mar 22 00:12:51 +1 Mar 22 00:26:55 hey trolls Mar 22 00:27:38 * _troll_ awakes Mar 22 00:30:10 Hmm. Should a touchscreen have its own SPI MISO/MOSI/SCK? Mar 22 00:30:19 I'm thinking yes. Mar 22 00:30:38 but then again I'm also thinking Arby's. Thoughts? Mar 22 00:31:10 n7segment: see what you did Mar 22 00:31:28 Shadyman: Arby's is good Mar 22 00:32:01 Shadyman: MISO/SCK make sense Mar 22 00:32:27 as i'd imagine a steady stream of touch data could be fairly high bandwidth Mar 22 00:34:34 Though, the GLCD itself runs on the 16-bit LCD bus, so spi would have touchscreen, maybe a flash memory ic, maybe SD Mar 22 00:35:13 I'd imagine SD can be used either on SD pins for parallelism, or on SPI with SPI mode Mar 22 00:36:05 I'm making a generic GLCD Cape Mar 22 00:38:16 Though really, maybe I should just make one for each pinout. Mar 22 00:38:51 it'd keep the EEPROMs cleaner/more accurate, and would let the LCDs be more centered Mar 22 00:51:11 what the efficency of a transformer Mar 22 00:51:58 i dont' even want to think about that Mar 22 00:54:03 really depends on the type ... fuel fired transformers transform faster ... the diesel ones often only become tow trucks Mar 22 00:55:06 ^ Mar 22 00:55:20 grrr. Mar 22 01:09:27 run with me here folks if you can Mar 22 01:09:40 i have a 200W PSu, that running at 85% efficence Mar 22 01:09:51 so, i have up to 30W of heat to get rid of Mar 22 01:09:55 i have a sealed box Mar 22 01:10:15 I cant' bolt the psu to the case, because of the way its running. Mar 22 01:10:25 if i put a small fan inside the box Mar 22 01:10:41 i can effectively make sure that there are no hotspots Mar 22 01:11:07 does anyone know how to calculate the thermal reistivey of a box? Mar 22 01:18:37 Does the beagleboard xm support dual displays? Mar 22 01:18:55 and how would that work? one display on the HDMI port and another out S-Video? Mar 22 01:20:09 All the youtube videos I've seen using the beagleboard to power video walls use one board per display. I've read that the board supports dual independent displays - so why not have one board for every two displays? Mar 22 01:20:28 why not use the beaglebone black Mar 22 01:20:29 :-) Mar 22 01:20:35 it has an HDMI port Mar 22 01:20:38 and its much cheaper Mar 22 01:20:45 or in fact a rasbperry PI Mar 22 01:20:48 its ery cheap Mar 22 01:23:03 would the bone or pi pack enough processing power to decode 720p streams, though? Mar 22 01:23:14 sure. Mar 22 01:23:16 the pi does Mar 22 01:23:24 the bone will easily Mar 22 01:23:47 I haven't seen much on using pis for setting up video walls, though Mar 22 01:24:13 I've seen beagleboard video walls done with omapfbplay, however Mar 22 01:25:22 we do not speak of r-pi here Mar 22 01:25:56 thanks mrpackethead, but as mranostay brings me back to my original question.... :-) Mar 22 01:26:15 mranostay: it was a troll trap i had set Mar 22 01:26:27 it's trap! Mar 22 01:26:36 you certainly can do video on the beagle Mar 22 01:26:49 on the bone Mar 22 01:27:04 right now you need a cape Mar 22 01:27:12 but soon the *new* secrect thing Mar 22 01:27:15 will be avaialbe Mar 22 01:27:19 and you won't Mar 22 01:27:30 it woudl be an ideal and low cost platform for large video wallss Mar 22 01:28:15 I don't follow you.. you're suggesting there's a model release pending that will be more ideal for large video walls? Mar 22 01:28:30 yes. Mar 22 01:28:43 any idea how soon might be soon? Mar 22 01:29:13 and what may make this secret thing more suited to larger video walls? Mar 22 01:34:42 mranostay: i have got a brillant job interview question for you. Mar 22 01:35:12 https://fbcdn-sphotos-g-a.akamaihd.net/hphotos-ak-ash3/525112_10152637009320577_264696585_n.jpg Mar 22 01:36:11 I guess both of my last questions delved too deep into the "secret" in "secret thing" Mar 22 01:36:17 http://www.element14.com/community/community/knode/dev_platforms_kits/element14_dev_kits/next-gen_beaglebone?CMP=KNC-APAC-Knode-NGBeagleB Mar 22 01:36:33 foozo, no just the people who know are not allowed to tell Mar 22 01:36:43 but theres *some* information in that link. Mar 22 01:37:07 http://beagleboard.org/unzipped <--- and some here Mar 22 01:37:20 well, that's much more informative, thanks :) Mar 22 01:38:05 "just in time for spring" is good enough for me :-) Mar 22 01:39:10 ah that one Mar 22 01:39:29 for delving into video walls, I originally asked about using the beagleboard xm vs. the bone simply due to the fact that the xm had hdmi output. The 4xUSB hub on board looked a bit more useful, too. Mar 22 01:39:37 well what you define just in time for spring? :) Mar 22 01:39:50 june 20th is still spring :) Mar 22 01:41:01 mranostay: my definition for that falls "anytime now". Usually "just in time for" some time period means before that time period hits. But since spring is officially here, that may not be applicable in this case, I guess. Mar 22 01:42:15 foozo: well it is just in time for me :P Mar 22 01:43:33 foozo: there will be plenty of grunt in it Mar 22 01:43:51 'grunt-included'(tm( Mar 22 01:44:51 grunt, referring to the vocal expression? indicating that it will pack a harder punch processor and feature wise? Mar 22 01:44:58 all this cryptic talk... :-p Mar 22 01:58:20 mrpackethead: this is actually quite exciting news.. i don't yet actually own a bb.. but bb ng looks very, very attractive. Mar 22 01:58:54 yes, it does look good Mar 22 01:58:55 :-_ Mar 22 01:59:53 i think I figured out how to still have a massive coin cell on the GPS/GSM verison of my GPS Cape... Vertical 2032 holder :3 Mar 22 02:00:06 and probably on the outside of the cape headers, so it's stackable Mar 22 02:00:20 http://www.digikey.ca/product-detail/en/BS-5/BS-5-ND/275304 tada Mar 22 02:04:44 Shadyman: where'd you get your gps cape? got a link? Mar 22 02:30:22 foozo: his gps is from the cheese factory Mar 22 02:30:29 swiss cheese factory Mar 22 02:34:56 yes, I realized Mar 22 02:34:57 so yummy Mar 22 02:35:24 mrpackethead: no, no, my protocape is from the swiss cheese factory. **** ENDING LOGGING AT Fri Mar 22 02:59:58 2013